From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cooper Subject: Re: Does Xen support nested VM migration Date: Wed, 1 Jul 2015 10:58:57 +0100 Message-ID: <5593B9E1.7080108@citrix.com> References: <1435718964.3287.17.camel@huaitong-MOBL.ccr.corp.intel.com> <55939BBF.6010300@citrix.com> <1435740948.5710.22.camel@huaitong-MOBL.ccr.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1435740948.5710.22.camel@huaitong-MOBL.ccr.corp.intel.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: "Han, Huaitong" Cc: "xen-devel@lists.xen.org" List-Id: xen-devel@lists.xenproject.org On 01/07/15 09:56, Han, Huaitong wrote: > On Wed, 2015-07-01 at 08:50 +0100, Andrew Cooper wrote: >> On 01/07/2015 03:49, Han, Huaitong wrote: >>> Hi, all, >>> >>> When I create a L1 guest with nestedhvm=1, and create a L2 guest in L1 >>> guest, then migrate L1 guest, but the operation fails, and I find the >>> bugfix requires a lot of work. >> Fails in what way? > The xen hypervisor in L1 guest has execute "vmxon" and > nvmx->vmxon_region_pa get the value, after L1 migration, vmxon impact is > lost and nvmx->vmxon_region_pa==0 , when L2 guest has some operation > like VMREAD or VMWRITE, vmx_inst_check_privilege will return 1, and L1 > guest maybe hang or panic. Ah yes - I remember this now. I raised it as a concern about a year ago when rewriting migration and finding no information along these lines in the migration stream. To support this, you will need to add a new per-vcpu record in the hypervisor migration stream which sends the VMX architectural state from the source to the destination. ~Andrew